jay123035
Posts: 32
Joined: Wed Nov 21, 2012 5:24 am

power supply circuit

Tue Jul 30, 2013 7:30 pm

I am trying to assemble the power supply circuit out of the October 2012 issue of MagPi, but am having the worst luck.

This is the first time i have used a lot of these complaints so i am a little concerned. (and my first 3 attempts have not worked)

i think i am getting the grounds wrong.

the first picture is what i think i need to do.
the second picture is from the magazine

am i correct, or am i making a noob mistake?
Attachments
MagPi power circuit 2.GIF
MagPi power circuit 2.GIF (24.41 KiB) Viewed 2354 times
MagPi power circuit.GIF
MagPi power circuit.GIF (21.03 KiB) Viewed 2354 times

User avatar
Tage
Posts: 287
Joined: Fri May 24, 2013 2:29 am
Location: St Thomas, Ontario Canada

Re: power supply circuit

Wed Jul 31, 2013 1:41 am

I think you got it right,
but ground is ground is ground. why do you have different signal names on Ground? if you connect it like it is shown in the schematics, the ground connection is Ground. nothing else.
Now, this does not mean that Ground has to be connected to earth ground, it is just a name for something that we call Ground. all other voltages are referenced to this node, to make things simpler.

if anybody needs more detailed explanation just send a private message. I don't mind sharing my knowledge of electronics. I have been working mostly in power electronics since 1977, and there is no question that is too stupid.

danjperron
Posts: 3502
Joined: Thu Dec 27, 2012 4:05 am
Location: Québec, Canada

Re: power supply circuit

Wed Jul 31, 2013 2:46 am

Hi Jay123035,

Both schematics are the same!

Except label error on the first schematic.
Both side should be label "ground" and not -12V and -5V. If you don't like the term "ground" use "0V".

What is exactly the problem with it? Do you have a voltmeter ? What is the voltage at the output?

When you test the unit please but a small load and not the Rpi! Just in case of failure.

Could you post a picture of your prototype with all things connected? This will help.


Daniel

User avatar
Burngate
Posts: 6290
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: power supply circuit

Wed Jul 31, 2013 9:11 am

Earth = Ground = Chassis = 0v = reference point in most cases

Calling that left-hand end of your ground wire "-12v" and the right-hand end "-5v" is somewhat like saying the top of Everest is at 30,000 ft and so the bottom of Everest is "-30,000 ft"

Voltage is really a sloppy misnomer - it should be PD or Potential difference. "Difference" is the important word - the potential difference between this point and somewhere else.

danjperron
Posts: 3502
Joined: Thu Dec 27, 2012 4:05 am
Location: Québec, Canada

Re: power supply circuit

Wed Jul 31, 2013 11:00 am

Hi Burngate,

More like telling that the sea level is at -5000m from the east side of Everest and the sea level is at -12000m from the west side. But the sea level, except for tide and earth rotation effect, is roughly the same everywhere.

If you use ground has your reference point , you should consider it with a voltage potential of 0 Volts, unless you are using floating ground but this is an other issue. Everything is relative to a reference point. In electronic it is about difference in potential (Nothing absolute).

N.B. The Mont Everest has a height of 8848 meters (29029 feet) above the sea level. This imply that sea level is the reference point. (You were not too far off with 30000 feet)


Still waiting for the picture….
Daniel

Ravenous
Posts: 1956
Joined: Fri Feb 24, 2012 1:01 pm
Location: UK

Re: power supply circuit

Wed Jul 31, 2013 1:27 pm

Come along chaps. It's obvious Jay's got the right circuit, he just misunderstood what earth/0V is. We don't need geodetics, at least we don't yet :)

If he comes back, we just need to determine how his power supply is wired to it (and whether it's properly DC!), whether he's using the right choke, whether he's damaged any of the semiconductors in earlier tests, whether there's a layout problem (I'm not a switched PSU expert so I wouldn't know about that), etc.

jay123035
Posts: 32
Joined: Wed Nov 21, 2012 5:24 am

Re: power supply circuit

Wed Jul 31, 2013 5:39 pm

Thank you all for your responses.

I have a multimeter.
I tested for 5v between the 100uF, 100uH, feedback junction and ground. (no voltage)
I then tested between the pin 2 and pin 3 of the LM2576t -5 with no voltage (12v between pin 1 and pin 3)

The 100uH has a white dot over one prong, does that mean something? (the data sheet didn’t say anything, that I could find. Google was no help ether)

is the multimeter enough of a load? if not what would be a good way to add a small load? LED? a small motor?

The leads on the 1n5822 would not fit into my breadboard so I used a protoboard and tried soldering things together. After 3 attempts it looks really bad. I will clean it up; make a 4th attempt and post results and pictures tonight. (PST)

At the moment my power is a small 12 battery.
My goal is to power the Pi from a standard 12v car electrical system

Thank you all for your help.

Ravenous
Posts: 1956
Joined: Fri Feb 24, 2012 1:01 pm
Location: UK

Re: power supply circuit

Thu Aug 01, 2013 8:20 am

Are you absolutely sure the earth line is really connected all the way through? (Some of the breadboards have long rails along the top and bottom, which aren't necessarily connected all the way - some models have a break in the line in the middle. Fooled me with some new breadboards I bought recently!)

I would measure the input voltage right at the chip (pin 1) - that should be the proper DC voltage, if not there might be a dodgy connection or a short circuit somewhere.

Make sure the wires from the chip output to the diode, coil and capacitor are reasonably short. There's supposed to be a high frequency oscillation going on there and it can be unstable if the wires are too long.

Make sure the feedback connection is coming from close to the capacitor, too.

Don't use an LED as a load, remember it needs a resistor (a few hundred ohms?) and must not be cnnected to 5V without it. An old style torch bulb will work if you have one, or a very small motor. But as you're measuring zero volts at the output, it might be something else.

I really don't know much about switch mode supplies (despite the length of my replies!) so I don't know what else to check for... usually I just buy low current ones off the shelf. Hope you get it going!

jay123035
Posts: 32
Joined: Wed Nov 21, 2012 5:24 am

Re: power supply circuit

Thu Aug 01, 2013 5:03 pm

made a 4th attempt last night. still not as clean as i would like.

after soldering i rapped the wires around the leads to help hold the component in place and to show what wire goes were. (though i am worried about vibrations so i might cover the bottom in epoxy when i am finished so that it moves as one thing.)

i did not get a chance to test it as it was getting late and i still need to connect the grounds and +12v to the terminal block. so i will update tonight.

fingers crossed
Attachments
LM2576T-5.jpg
LM2576T-5.jpg (28.82 KiB) Viewed 2103 times
front.jpg
front.jpg (34.94 KiB) Viewed 2103 times
back2.jpg
back2.jpg (35.79 KiB) Viewed 2103 times

danjperron
Posts: 3502
Joined: Thu Dec 27, 2012 4:05 am
Location: Québec, Canada

Re: power supply circuit

Thu Aug 01, 2013 8:55 pm

Hi Jay,


just a quick look at your picture and I found that your diode is wrongly connected !

The cathode is on ground (white wire).

It is hard to decipher with your wire looping in the hole of the protoboard.

If your diode is inverted like I think it is , it is possible that the diode ant the I.C. is burned.

The cathode on diode is identified by a paint collar, in you case it is white.

Daniel
Last edited by danjperron on Fri Aug 02, 2013 12:48 am, edited 1 time in total.

jay123035
Posts: 32
Joined: Wed Nov 21, 2012 5:24 am

Re: power supply circuit

Thu Aug 01, 2013 10:13 pm

is there anyway to test the diode and the I.C.?

danjperron
Posts: 3502
Joined: Thu Dec 27, 2012 4:05 am
Location: Québec, Canada

Re: power supply circuit

Thu Aug 01, 2013 10:30 pm

Yes to test the diode it is easy,

Some multimeters have a diode test mode check if your multimeter has it.

If not use a 1K ohm resistor in serie with the diode to 5V supply.

The diode will be ok when the voltage on anode vs cathode is around 0.7V and if you flip the diode the voltage should be -5V.
(Voltage on the diode between the anode and cathode)

And for the I.C. it will be more tricky, maybe just a 1K home resistor at the output without diode and coil. and check if the voltage is around 5V.

B.T.W. If you just flip the diode , assuming that you find the diode ok, does it work?

Daniel

jay123035
Posts: 32
Joined: Wed Nov 21, 2012 5:24 am

Re: power supply circuit

Tue Aug 06, 2013 2:43 am

i am getting 88.7mV between ground and feedback/output. idk ill test the diode and the IC and go from there.

it cant be this hard. it was in MagPi

Return to “Automation, sensing and robotics”